My name is Brian Larson

1
There are two more proposals here than in my abstract.

My brief talk will take the following route:
1.  First, I’ll discuss previous studies that used gender as a variable. This
justifies me bringing this topic before you today and illustrates the problems
I’ve seen.
2.  Second, I’ll address the source of my discomfort.
3.  Third, I’ll discuss each of my proposals briefly.
4.  Finally, I’ll describe an empirical study I conducted that addressed these
concerns

2
3
4
So, for example, in some older studies in TC:
•  Sterkel (1988): Statistical analysis of writing by students looking for
differences correlated to gender.
•  MLS (1989): Study of how students interact in technical communication
class.
•  Tebeaux (1990): Looked at differences in quality of writing based on author
gender.
•  Allen (1994) considered whether women writing theoretical articles in tech
comm journals exhibited characteristics commonly attributed to authoritative
writing.

In some more recent studies in prof and TC:
•  Wolfe and Powell (2006) looked at complaining by students in collaborative
work environments in TC education to see whether practices differed based
on gender.
•  Wolfe and Powell (2009) looked at student perceptions of verbal interaction
strategies previously associated with men or women in previous studies.
•  Plumlee et al. (2016) looked at writing prepared by students and assessed
gender differences/similarities with regard to readability, I vs. you
orientation, and tone. No explanation of the ascription of gender to the
students’ writing.

5
1.  They don’t say what gender is. In other words, they don’t say what they
mean by “gender,” or by “men and women” or by “male and female” etc.
2.  They don’t say how they determined the gender of the participants. In other
words, we don’t know how the researcher figured out who were the Gender
M and who were the Gender F participants in the study.

6
Studies in computational and corpus linguistics have also looked at gender as
a central concern. For example,
•  Argamon et al (2003) and Koppel, Argamon & Shimoni (2002) performed a
computational analysis of texts from the British National Corpus to see
whether there were differences based on author gender. The BNC is a
collection of published texts; the researchers used (without any comment)
the gender labels applied to authors of BNC texts by third parties.
•  Argamon et al (2007) and Yan and Yan (2006) used blog profile account
settings.

So there we know how they assigned the gender labels. But things sometimes
get more interesting:
•  Herring and Paolillo (2006) looked for differences in linguistic characteristics
in blog posts by women and those by men. The assigned gender to blog
authors “by examining each blog qualitatively for indications of gender such
as first names, nicknames, explicit gender statements... and gender-
indexical language.” This is a sort of heuristic approach.
•  Rao et al. used a computational heuristic: “For gender, the seed set for the
crawl came from initial sources including sororities, fraternities, and male
and fe- male hygiene products. This produced around 500 users in each
class.”

7
8
9
10
11
12
Self-iden=fica=on is probably the gold standard. Except in circumstances where you
might not expect complete candor, you can count on a par=cipant to say what her
own gender is.

Keep in mind in some cases you might want to know what others think. For example,
if studying whether a teacher treats students differently based on student genders,
you want to know what genders the teacher ascribes to students.

13
Respect difficul-es of research par-cipants when asking them to self-iden-fy for
gender.

In contemporary American college classrooms, it’s not unusual to have students who
do not easily iden=fy with one gender or another or who ac=vely refuse to be classed
in a par=cular gender. Others are confidently transgendered.

There are so many ways that folks might choose to describe their genders that lis=ng
them might also be imprac=cal, especially as the list itself might have reac=ve effects

14
15
I’m using methods of corpus and computational linguistics to assess whether
the writing of Gender M and Gender F authors exhibits differences tied to their
genders.

Definition: Gender knowledge is knowledge about how to signal one’s sex to
other humans in appropriate circumstances.

Gender, in sexual animals, is thus an adaptive characteristic, tied with sex,
reproduction, and survival. Consequently, gender categories are of such
pervasive importance to members of human societies that it’s difficult to
interact without reference to them.

16
17
18
My readers have the choice whether to accept my argument or to challenge it.

But ul=mately, I’ve made my decision-making transparent and subject to inquiry
while respec=ng my par=cipants.

19
20

Master your semester with Scribd & The New York Times

Special offer for students: Only $4.99/month.

Master your semester with Scribd & The New York Times

Cancel anytime.